Cool! I've only looked in a browser so far, and it looks good. All that it really needs to be interpretable in the RDF model is to identify important nodes with URIs - "Open text #1" will work on your site in a browser, after a URI has been entered, so I suspect you've got 99.8% in place already. (Ok, I've not looked at the source yet ;-)
How it helps is that it makes it unambiguous in the global Web thing, so (potentially) everyone can use a uniform approach to using the stuff.

(But please ping me and/or Shelley again)

hey, i've been researching RDF as well, and have the exact frustrations as u..however i stumbled across a tool that I may use to get my Database out to the web its called D2RQ it is located
http://www.wiwiss.fu-berlin.de/suhl/bizer/d2rq/spec/

check out the manual, it uses mappign files and your database schema to generate RDF support for Seasme, and u would publish the RDF using http://www.joseki.org/ or seasme...

personally they really need more tools to get legacy data to the web D2RQ is a step in the right direction..somthing cool would be (middlegen->D2RQ mapping), middlegen already generates Hibernate,EJB files..

Jeryl
http://pharaohofkush.blogspot.com/